Customers have never enjoyed such high levels of convenience and connection, and nor have fraudsters, who are quick to adapt their tools and techniques to exploit vulnerabilities in the financial ecosystem.
Sebastian Takle, Head of Financial Crime Center at DNB on the wave of digital-enabled fraud
‘In 2021-2022, APP fraud overtook card fraud in both value and volume, and this switch was a real awakening about the new direction fraud was taking,’ said Michael Hammond, Fraud Product Owner at Pay.UK, adding that account takeover fraud has decreased as a result of new security measures by the banks.
‘ATO fraud is no longer so profitable for fraudsters, so they’ve switched to social engineering scams instead because it’s easier to target the individual than the technology.’
How is the face of the fraudster changing? Sebastian Takle, Head of Financial Crime Center at DNB tells us more:
Financial institutions are having to work harder and smarter to combat these types of sophisticated scams, which can rely on advanced technologies and psychological grooming techniques. Generative AI is increasingly seeing use – condensing fraud alerts so swamped reviewers can quickly assess those that need further investigation or delivering entity insights to work out which customers may be most vulnerable to different types of scams.
Aaron Elliot-Gross, Global Director of Product Compliance at Wise, highlighted the importance of predictive machine learning models in the face of changing, and more sophisticated fraudulent behaviours.
Machine learning models look at customer onboarding and transaction behaviours to see if it’s a potential bad actor or not. Often it can be someone with a real identity, which has then been immediately sold to a criminal network. Multiple accounts where the selfie for the ID has been taken with the same background – in the same car, for example – is a giveaway.
AARON ELLIOT-GROSS
GLOBAL DIRECTOR OF PRODUCT COMPLIANCE, WISE
Biometrics, such as facial or voice recognition, can help by confirming identity or ownership of an asset but there was more ambivalence about the value of behavioural analytics, which flags when a customer may be using the app erratically and is possibly under duress.
‘It sounds good but I suspect there are lot of false positives because customers can be erratic from time to time,’ said Aaron Elliot-Gross of Wise.
Sebastian Takle, Head of Financial Crime Center at DNB, said biometrics can enable quick and frequent rechecks of identity without disrupting the customer experience to make sure, for example, the person who started an online banking session is still the same person now they’re authorising a payment transfer.
He highlighted the inherent tension between delivering seamless CX that meets the customer’s expectation of convenience, speed and control with a bank’s consumer duty and anti-fraud obligations.
‘If we have to pause or cancel a transaction because we think it might be fraud, then that’s absolutely what we should do,’ Takle said. ‘It doesn’t have to be a bad experience for the customer, it depends on how you handle that. But if you’re in any doubt, you do not let that transaction happen.’
All models, of course, rely on good data to be effective and that’s where collaboration is key in order to share information to identify emerging fraud trends. This can be informal, or via more formalised consortium-based models.
‘As the PSO, we are looking at a centralised data store of Faster Payments data, with analytics sitting on top, to see if looking across an entire network improves detection outcomes,’ said Michael Hammond at Pay.UK.
As always, check back with MoneyLIVE for updates on new initiatives in the fight against fraud.
Stay in the loop with MoneyLIVE
Subscribe to our newsletter to receive news, insights and special offers.